A novel knowledge-aided reduced rank (RR) space time adaptive processing (STAP) method is proposed in this paper. Firstly, a clutter model was constructed by a priori knowledge such as flying parameters, radar system parameters, mapping and terrain parameters. Secondly, the clutter was performed via the pre-whiten processing of the constructed clutter model. Finally, RR STAP method was used to suppress the residual clutter. The simulation results show the KA-RR method is more robust in nonhomogeneous clutter environment compared with conventional RR STAP method.